Analysis of the Gaia Data Release 3 parallax bias at bright magnitudes

Abstract: The combination of visual and spectroscopic orbits in binary systems enables precise distance measurements without additional assumptions, making them ideal for examining the parallax zero-point offset (PZPO) at bright magnitudes (G < 13) in Gaia. We compiled 249 orbital parallaxes from 246 binary systems and used Markov Chain Monte Carlo (MCMC) simulations to exclude binaries where orbital motion significantly impacts parallaxes. After removing systems with substantial parallax errors, large discrepancies between orbital and Gaia parallaxes, and selecting systems with orbital periods under 100 days, a final sample of 44 binaries was retained.The weighted mean PZPO for this sample is -38.9 $\pm$ 10.3 $\mu$as, compared to -58.0 $\pm$ 10.1 $\mu$as for the remaining systems, suggesting that orbital motion significantly affects parallax measurements. These formal uncertainties of the PZPO appear to be underestimated by a factor of approximately 2.0. For bright stars with independent trigonometric parallaxes from VLBI and HST, the weighted mean PZPOs are -14.8 $\pm$ 10.6 and -31.9 $\pm$ 14.1 $\mu$as, respectively. Stars with $G \leq 8$ exhibit a more pronounced parallax bias, with some targets showing unusually large deviations, likely due to systematic calibration errors in Gaia for bright stars. The orbital parallaxes dataset compiled in this work serves as a vital resource for validating parallaxes in future Gaia data releases.